One eventual goal of genetics-based therapies involves the transfer of nucleic acids to bone marrow repopulating cells and the stable expression of the encoded gene over the lifetime of the patient. This approach may eventually be useful for the treatment of prevalent genetic diseases including the thalassemias and sickle-cell anemia. However, before gene therapy can be fully realized, basic paradigms involving the transfer of nucleic acids and their regulated expression in mammalian cells must be defined. The stable expression of transferred nucleic acids in mammalian cells remains a poorly understood phenomenon. We have begun the study of stable green fluorescent protein (GFP) gene expression in mammalian cells. GFP has been used successfully as a reporter molecule for transient gene expression, but reports of stable GFP expression are sparse. Thus, the objective of our work was to understand the conditions required to produce cell lines that exhibit durable, high-level GFP expression. Chinese hamster ovary (CHO) cells were transfected with plasmid DNA encoding both GFP and neomycin phosphotransferase (neo) cassettes. Flow cytometric and visual monitoring of the cells cultured in the presence and the absence of G418-mediated selective pressure revealed stable expression in pools after more than 12 weeks in nonselective medium. This observation suggests that GFP itself does not bestow a growth disadvantage in mammalian cells. Our goal is to produce hematopoietic cell lines and primary hematopoietic cells that stably express GFP in the absence of selective pressure. Once identified, these cells will permit further study of stable gene expression.